[Eeglablist] How can I reject artifact in CNT based on a criteria specific to one or multiple electrodes?

[alvin kao]

When M1 or M2 has been contaminated (i.e. extremely high or low amplitude
as can be easily seen in CNT file), I found the entire EEG data will be
contaminated if I conducted a re-reference by average of M1 and M2.
Artifacts are introduced by abnormal M1 or M2 and then ruin the subsequent
ERP waveform.

I tried to fix this by manually rejecting CNT blocks that contain evident
artifacts on M1 and M2. By doing so, the subsequent ERP waveform looks much
better. However, this manual rejection takes me a lot of time as I need to
check the entire CNT file by hand. Also, it is a very subjective process to
decide if a CNT block should be rejected or not.

To be efficient and consistent, my idea is to set a +/- threshold for both
M1 and M2. If the amplitude of M1 and M2 in certain period exceed the
threshold then that period will be rejected and would not be included in
the subsequent re-reference.

However, in reality I do not know how to do this either by using GUI or
script.
Please advise if anything I can try.

Thanks!

Alvin Kao
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://sccn.ucsd.edu/pipermail/eeglablist/attachments/20140617/9c551d0c/attachment.html>